Chapter 8 Frostbite
Frostbite is a true freezing injury, and depending on the severity, ice crystals will tend to form in deep and/or superficial tissues. The degree of injury is a complex interaction of many factors, including environmental (e.g., temperature, windchill, length of exposure, altitude) and individual (e.g., genetics, comorbidities, medications and drugs, clothing, skin products, previous frostbite injury, among other factors).
This chapter reviews current understanding of the history, epidemiology, physiology, pathophysiology, classification, risk factors, clinical presentation, field and hospital evaluation and treatment, prevention, and consultation strategies for the management of frostbite. Although the principles of prevention and treatment of frostbite transfer from urban to wilderness settings, the time to definitive treatment is often prolonged and resources are almost always limited in the wilderness setting; therefore, wherever possible, we incorporate special considerations for wilderness management of frostbite injuries. Hypothermia and nonfreezing cold-induced injuries are discussed in Chapters 5 and 7.
A 5000-year-old pre-Columbian mummy discovered in the Chilean mountains is widely recognized as the earliest documented evidence of frostbite.138 In 218 BC, Hannibal lost nearly one-half of his army of 46,000 to cold injuries in only 15 days when they crossed the Pyrenean Alps. Dr. James Thatcher wrote in 1778 that 10% of George Washington’s army had been left to perish in the winter cold during his campaign against the British soldiers.151
Despite the fact that frostbite was a known consequence of military campaigns in the cold for thousands of years, the first authoritative report of mass casualties was by Baron Larrey,93 surgeon-in-chief of Napoleon’s army during the invasion of Russia in the winter of 1812-1813. Larrey introduced the concept of friction massage with ice or snow, avoidance of heat in thawing, and the idea that cold injuries were similar to burn injuries. These concepts are better understood against the background as viewed by Larrey;93 soldiers with cold injuries rapidly rewarmed their extremities over roaring fires (65° to 75° C [149° to 167° F]) after long marches, only to renew the trek and refreeze their extremities the next day. Larrey recognized that warming was good, but he cautioned against the use of excessive heat and ultimately recognized the freeze–thaw–freeze cycle. Napoleon left France with 250,000 men and returned 6 months later with only 350 effective soldiers. The remainder were casualties to cold or starvation.93
During both world wars and the Korean conflict, at least 1 million cases of frostbite occurred.129,130,167,184 High-altitude frostbite, first described in 1943, was recognized from the treatment of aviators in World War II, when gunners of aircraft flying between 7620 and 10,668 m (25,000 and 35,000 feet) fired machine guns through open waist pots, removing their bulky mittens and jackets to improve the dexterity that they felt was crucial to saving their lives.184
Until the 1950s, treatment of cold injuries basically followed Larrey’s guidelines. In 1956, experimental laboratory work encouraged Merryman, the Public Health Service district medical officer in Tanana, Alaska, to try rapid rewarming at 37.8°C (100° F) on a patient with frostbite and hypothermia.20,117 This was the genesis and has become the cornerstone of the method currently used in Alaska and popularized by Mills.117,119
No comprehensive statistical data are available on the incidence of frostbite, but it is much more prevalent during military campaigns and is a known hazard for mountain climbers and polar explorers. The typical frostbite victim is working or recreating in cold and/or high-altitude environments, is homeless, or is accidentally trapped outdoors in the winter;86,89,133 is 30 to 49 years old;136,181 and in more than half of urban cases, is intoxicated.115,177
In the nonadventurer U.S. civilian population, Mills had collected 500 cases in Alaska by 1963, Cook County Hospital in Chicago recorded 843 cases from 1962 to 1972, and Detroit Receiving Hospital reported on 154 patients treated between 1982 and 1985.12,64,90,117 The incidence of frostbite in Finland, calculated in a 9-year retrospective query of hospital admissions, was 2.5 cases per 100,000 inhabitants.76
The mean annual incidence of frostbite in the Finnish military was reported as 1.8 per 1000, with the head, hands, and feet most commonly affected.95 In a more recent study of almost 6000 Finnish military recruits, 44% reported at least one episode of frostbite during their lifetime, and the annual incidence was 2.2%; the head, hands, and feet were the sites most commonly afflicted.38 Among refugees navigating a high-altitude military line of control during December 1988 to March 2003, 2564 cases of frostbite were treated at local hospitals in Muzaffarabad, Azad Kashmir.81
A 10-year retrospective survey of medical records of the British Antarctic Survey revealed that the incidence of frostbite was approximately 65.6 per 1000 per year.23 During a 10-year period in the Karakorum mountains, 1500 cases of frostbite were treated at tertiary care medical facilities; all victims were males ages 17 to 43 years. The incidence is unknown because the total number of potential exposures was unrecorded.61 In a questionnaire-based study of 637 mountaineers, the mean incidence of frostbite was 366 per 1000 population per year,60 and out of 2219 south-side Mt Everest climbers in 7 years (2001 to 2007), the base camp medical clinic at Mt Everest saw 35 cases of frostbite (and estimates at least a comparable number did not seek treatment at the facility).47
Physiologically, humans are tropical beings, better suited to losing heat than retaining it. When naked and at rest, a person’s neutral environmental temperature is 28° C (82.4° F); with an environmental drop of only 8° C (14.4° F), metabolic rate must double to avoid lowering of body temperature.
The rate at which heat is lost by radiation is a function of the temperature of the cutaneous surface, which in turn is primarily a function of the rate of blood flow through the skin. Heat is poorly conducted from warmer internal tissue to the cutaneous surface because adipose tissue is a good heat insulator.
As a result, the cutaneous circulation is key to the development of frostbite. Because of its role in thermoregulation, normal blood flow of skin far exceeds its nutritional obligation. The skin holds a complex system of capillary loops that empty into a large subcapillary venous plexus containing the majority of the cutaneous blood volume. Under normothermic conditions, 80% of an extremity’s blood volume is in the veins of skin and muscle. Skin blood volume depends in part on tone in resistance and capacitance vessels, and tone in turn depends largely on ambient and body temperature. Under basal conditions, a 70-kg (154.3-lb) person has a total cutaneous blood flow of 200 to 500 mL/min. With external heating to maintain skin temperature at 41° C (105.8° F), this may increase to 7000 to 8000 mL/min, whereas cooling the skin to 14° C (57.2° F) may diminish it to 20 to 50 mL/min.
Blood flow through apical structures such as the nose, ears, hands, and feet varies most markedly because of the richly innervated arteriovenous connections. Blood flow to hand skin can be increased from a basal rate of 3 to 10 mL/min/100 g of tissue to a maximum of 180 mL/min/100 g of tissue. This cutaneous vascular tone is controlled by both direct local and reflex effects. Indirect heating (warming a distant part of the body) results in reflex-mediated cutaneous vasodilation, whereas direct warming results in vasodilation dominated by local effects. When both types (central and peripheral) of heating or cooling are present, their effects are additive.
Cutaneous vessels are controlled by sympathetic adrenergic vasoconstrictor fibers, and vascular smooth muscles have both α- and β-receptors. Vasodilation in the hands and feet is passive, so maximal reflex vasodilation occurs after sympathectomy.
When the hand or foot is cooled to 15° C (59° F), maximal vasoconstriction and minimal blood flow occur. If cooling continues to 10° C (50° F), vasoconstriction is interrupted by periods of vasodilation and an associated increase in blood and heat flow. This cold-induced vasodilation (CIVD), or “hunting response,” recurs in 5- to 10-minute cycles to provide some protection from the cold. There is considerable individual variation in the amount of CIVD, and it is believed this might explain some of the variation in susceptibility to frostbite. Prolonged repeated exposure to cold increases CIVD and offers some degree of acclimatization. Inuit, Sami, and Nordic fishermen have a strong CIVD response and very short intervals between dilations, which may contribute to maintenance of hand function in the cold environment.57
Normal skin maturation and tissue function rely on maintenance of permeability and integrity of all tissue membranes. A steady-state relationship of prostaglandins (PGs), particularly PGE2 (vasodilator) and PGF2α (vasoconstrictor), is crucial for normal skin function. An imbalance may disrupt cell membrane equilibrium. This relationship is controlled through PGE2-9-ketoreductase and nicotinamide adenine dinucleotide phosphate (NADPH). Low concentrations of PGE2-9-ketoreductase found in normal skin emphasize an active biologic presence.
It is not difficult to imagine why frostbite tends to affect tissues that are acral and receive diminished blood supply as a result of vasoconstriction (e.g., fingers, toes, ears, nose), thereby conserving heat for the core. The nose and corneas may be difficult to protect from cold wind and are particularly vulnerable; face coverings such as a balaclava and goggles should be considered in extreme conditions. In addition, men who jog, ski, or otherwise exercise in the cold may be prone to penile frostbite, especially if fast speeds create a headwind. Clothing then becomes moist with sweat in the groin, and the tip of the penis is vulnerable, especially if unprotected by a wind-resistant or windproof outer layer.47,66
The pathophysiology of frostbite has been categorized in several different ways, illustrating the numbers of variables that affect the extent and depth of tissue damage. Frostbite may be divided into four pathologic phases: prefreeze, freeze–thaw, vascular stasis, and late ischemic.
Overlap occurs among these phases. The changes during each phase vary with rapidity of freezing and duration and extent of injury. Some feel that it is conceptually clearer to divide pathologic changes occurring in frostbite into two categories: those resulting from direct cellular injury and those from indirect cellular effects, or progressive dermal ischemia,125 a similar pathophysiology seen in thermal burn patients.109,110,150
Cells subjected to a slow rate of cooling (hours) develop ice crystals extracellularly in the cellular interspaces. Rapid cooling (seconds to minutes) produces intracellular ice crystals, which are more lethal to the cell and less favorable for cell survival. In a clinical cold injury, the slower rate of freezing does not produce intracellular crystals;112,113 however, the extracellular ice formed is not innocuous. It draws water across the cell membrane, contributing to intracellular dehydration. The theory of cellular dehydration was originally proposed by Moran124 in 1929 and subsequently supported by Merryman’s study of “ice-crystal nucleation.”112,113,115 Cellular dehydration produces modification of protein structure by high electrolyte concentration, alteration of membrane lipids, alteration of cellular pH, and imbalance of chemical activity.106,107,116 This phenomenon subsequently permits a marked and toxic increase of electrolytes within the cell, leading to partial shrinkage and collapse of its vital cell membrane. These events are incompatible with cell survival.
Not all the water within a cell is freezable. A small amount of unfrozen water, “bound water,” constitutes up to 10% of the total water content and is held tightly in the protein complex within the cell. No matter how rapid or marked the cold injury, this bound water remains liquid. At temperatures below −20° C (−4° F), approximately 90% of available water is frozen.194 Although the theory of ice crystal disruption of cell structure is attractive, it has yet to be conclusively proved.
Thermal shock defines the phenomenon of sudden and profound temperature change in a biologic system. Precipitous chilling has been theorized to be incompatible with life, but the severity of this phenomenon is debatable. Another poorly understood concept is the manner in which subzero temperatures produce denaturation of lipid–protein complexes. One proposed theory hypothesizes the detachment of lipids and lipid protein from cell membranes as a consequence of the solvent action of a toxic electrolyte concentration within a cell.99,100 There is no direct evidence supporting an alteration of enzyme activity during freezing, but DNA synthesis is inhibited.74 On the other hand, there is indirect evidence of ox liver catalase inactivation caused by denaturation and structural alteration of lactic acid dehydrogenase after freezing and thawing.102,166
Indirect cellular damage secondary to progressive microvascular insults is more severe than is the direct cellular effect. This is supported by the observation that skin tissue subjected to a standard freeze–thaw injury, which consistently produced necrosis in vivo, survived as a full-thickness skin graft when transplanted to an uninjured recipient site.190 Conversely, uninjured full-thickness skin did not survive when transferred to a recipient area pretreated with the same freezing injury. Thus, direct skin injury is reversible. The progressive nature of injury is probably secondary to microvascular changes.
Approximately 60% of skin capillary circulation ceases in the temperature range of 3° to 11° C (37.4° to 51.8° F), whereas 35% and 40% of blood flow ceases in arterioles and venules, respectively.148 Capillary patency is initially restored in thawed tissue, but blood flow declines 3 to 5 minutes later. Three nearly simultaneous phenomena occur after thawing: momentary and initial vasoconstriction of arterioles and venules, resumption of capillary circulation and blood flow, and showers of emboli coursing through microvessels.197 Ultimately, there is progressive tissue loss caused by progressive thrombosis and hypoxia. This is similar to the tissue loss seen in the distal dying random flap and the no-reflow phenomenon. For both of these, in addition to the effect of arachidonic acid metabolites, oxygen free radicals have been shown to be detrimental and to contribute to tissue loss. It has been proposed that this may be the case with frostbite injury.19 Emerging concepts, perhaps including the deleterious role of protein kinase C in reperfusion injury, will no doubt be incorporated into our understanding of frostbite injury as scientific understanding evolves.
Considerable evidence points to the primary alteration of the cold injury being injury to vascular endothelium.196 Seventy-two hours after a freeze–thaw injury, there is loss of vascular endothelium in the capillary walls, accompanied by significant fibrin deposition. The endothelium may be totally destroyed, and fibrin may saturate the arteriole walls.122,196,197 Ultrastructural derangement of endothelial cells after the thaw period has been observed by electron microscopy in capillaries of the hamster cheek pouch following subzero temperatures.144 The endothelial injury was confirmed by demonstrating fluid extravasation from vessels almost immediately after thawing.197 As in other forms of trauma, vascular endothelial cells swell and protrude inward into the lumen until they lyse.
Venules appear more sensitive to cold injury than do other vascular structures, partly because of lower flow rates. Arterioles, with a rate of flow almost twice that of venules, are less damaged by freezing and develop stasis later than do venules. Capillaries manifest the fewest direct effects of cold injury, but their flow is quickly arrested as a result of their position between arterioles and venules. Generalized stasis and cessation of flow are noted at the point of injury within 20 minutes after freeze and thaw. “White thrombi” (blood cells and fibrin) follow platelet thrombi as blood flow progressively slows. Sludging and stasis result in ultimate thrombosis. Microangiography after cold injury shows that although spasm of the arterioles and venules exists, it is not marked enough to completely account for the decreased flow of progressive microvascular collapse.9 In the 1950s, Kulka87,88 observed that vascular thrombosis after cold injury advanced from the capillary level to that of the large vessels and ultimately resulted in ischemic death of progressively larger areas. Viable dermal cells may be observed histologically in cold-injured tissues for up to 8 days or until there occurs occlusion of local vessels. This emphasizes that a major role is played by vascular insufficiency and that direct injury to cellular structures and mechanisms may be reversible.
Because Cohnheim had shown changes in cold injury to be similar to changes seen in other inflammatory states, Robson and Heggers150 postulated that the progressive ischemia seen in frostbite might be caused by the same inflammatory mediators responsible for progressive dermal ischemia in the burn wound. They evaluated blister fluid from patients with hand frostbite, measuring levels of PGE2, PGF2α, and thromboxane B2 (TXB2). Levels of the vasoconstricting, platelet-aggregating, and leukocyte-sticking prostanoids (PGF2α and thromboxane A2 [TXA2]) were markedly elevated. The investigators postulated that massive edema after cold injury was due either to leakage of proteins caused by release of these prostanoids or to leukocyte sludging in the capillaries and increased hydrostatic pressure. Studies have confirmed the similarity between cold injury and the burn wound.19
Severe endothelial damage was observed by researchers studying a minimal cold-injury model in the hairless mouse.13 In addition, the sequence of endothelial damage, vascular dilation, vascular incompetence, and erythrocyte extravasation was confirmed. This led to speculation that arachidonic acid metabolites, which may originate from severely damaged endothelial cells, are important in progressive tissue loss. Significantly absent from in vivo and microscopic observations were vascular spasm, thrombosis, and fibrin deposition, all of which have previously been implicated as pathophysiologic mechanisms. A rabbit ear model demonstrated increased tissue survival after blockade of the arachidonic acid cascade at all levels.145 The most marked tissue salvage resulted when specific TXA2 inhibitors were used. This has now been shown to be effective in clinical situations.64
Reports in the 1940s documenting the histopathology of frostbite injury to the skin were not comprehensive. Historically, studies by several investigators have been limited to skin biopsies without documentation of location, exposure time, temperature, or time elapsed since the injury.159
More recently, experimental studies have been able to document the histopathology of skin changes under controlled conditions. In 1988, Schoning and Hamlet161,162 used a Hanford miniature swine model for frostbite injury (−75° C [−103° F] exposure for up to 20 minutes) to note progressive epithelial damage. Early changes included vacuolization of keratinocytes, and there was loss of intercellular attachments and pyknosis over a period of 1 week or more. This subsequently progressed to advanced cellular degeneration and formation of microabscesses at the dermoepidermal junction. Later changes included epithelial necrosis and regeneration, either separately or together within the same tissue. Such histopathologic data favor the current standard of conservative delayed-surgical management of frostbite injury.
However, Marzella and associates104 used a New Zealand white rabbit ear model of frostbite injury and proposed that the skin necrosis induced by frostbite injury was merely a reflection of damage to the target cell—the endothelial cell. After submersion of a shaved rabbit ear in 60% ethyl alcohol at −21° C (−5.8° F) for 60 seconds, the entire microvasculature demonstrated endothelial damage within 1 hour; erythrocyte extravasation occurred within 6 hours. These early vascular changes in the rabbit ear model are in contradistinction to the timing of vascular changes in the Hanford miniature swine model reported by Schoning and Hamlet.162 These clinicians performed biopsies on animals exposed to frostbite injury (−75° C [−103° F] for up to 20 minutes) and evaluated the specimens for vascular inflammation, medial degeneration, and thrombosis. The earliest change documented both grossly and microscopically was hyperemia. Within 6 to 24 hours, leukocyte migration and vasculitis were noted. However, the most severe vascular changes of thrombosis and medial degeneration were not observed for 1 to 2 weeks after the injury.
Whether or not changes in the epidermis are primary or secondary to damage of underlying endothelial cells, it is clear that these tissues have potential, although limited, capacity for regeneration. Human experience clearly suggests that robust local tissue inflammation and coagulation stimulate microvascular thrombosis and progressive cell death.125 A perfect representative animal model for frostbite has yet to be found. A wide range of animal models is used to create and assess the condition. Developing a consistent reproducible and appropriate model would facilitate frostbite research.171 As early as 1991, Manson and co-workers101 proposed that frostbite is characterized by acute tissue injury induced by freezing and thawing. Initial complete ischemia is followed by reperfusion and later, tissue necrosis. They suggested that the vascular events supported the hypothesis that free radical–mediated reperfusion injury at thawing might contribute to tissue necrosis after frostbite in a manner similar to that seen after normothermic ischemia. Supporting evidence included electron micrographs showing the appearance of severe endothelial cell injury beginning during freezing and extending through early reperfusion. Later, neutrophil adhesion, erythrocyte aggregation, and microvascular stasis were seen.
Classically, frostbite has been described by its clinical presentation, but this can be difficult to predict in the field and before rewarming.85,117 Mills124,126 favors the use of two simple classifications: mild (without tissue loss) and severe (with tissue loss.) Historically, and following the classification of thermal burn injury, frostbite has been divided into “degrees” of injury based on acute physical findings after freezing and rewarming.
Frostnip (Figure 8-1, online; also see Figure 9-8) is superficial and associated with intense vasoconstriction. It is characterized by discomfort in the involved parts. Symptoms usually resolve spontaneously within 30 minutes, and no tissue is lost. There is some question whether this qualifies as cold-induced injury, because neither frozen extracellular water nor progressive tissue loss is routinely demonstrated.
First-degree injury shows numbness and erythema. There may initially be a white or yellowish, firm plaque in the area of injury. There is no tissue loss, although edema is common (Figure 8-2). Second-degree injury results in superficial skin vesiculation (Figure 8-3). Clear or milky fluid is present in the blisters, surrounded by erythema and edema. Third-degree injury shows deeper blisters, characterized by purple, blood-containing fluid (Figure 8-4). This indicates that the injury has extended into the reticular dermis and beneath the dermal vascular plexus. Fourth-degree injury is completely through the dermis, and involves relatively avascular subcuticular tissues (Figure 8-5). This tends to cause mummification, with muscle and bone involvement (Figure 8-6). Less severe bone injury in children may affect the growth plate and result in developmental digital deformities.16,22
FIGURE 8-2 A Nordic skier with first-degree frostbite (central pallor having cleared after rewarming) of the abdominal skin. This skier reported having skied for 90 minutes in −23.3° C (−10° F) temperature, unaware that his shirt, underneath a parka, had come untucked from his trousers.
(Courtesy Luanne Freer, MD.)
FIGURE 8-3 A climber with second-degree frostbite of the fifth finger sustained after only several seconds’ exposure to −45.6° C (−50° F) windchill when gloves were briefly removed to handle placement of a carabiner to the fixed rope. Clear bullae developed after rewarming.
(Courtesy Luanne Freer, MD.)
FIGURE 8-4 A climber with second-, third-, and fourth-degree frostbite of the hand. Note fingers 1 to 4 with hemorrhagic bullae over the areas of third-degree injury, clear bullae over the dorsum of the hand with second-degree injury, and deeply violaceous and unblistered fourth-degree injury of the distal phalanx of the fifth finger.
(Courtesy Luanne Freer, MD.)
(Courtesy Luanne Freer, MD.)
Cauchy and colleagues26 proposed a different classification of frostbite injury for the hand and foot that is based on the risk for amputation of the affected part. An early prognosis for frostbite patients may be delayed by the lack of useful clinical guidelines; this new classification scheme (Tables 8-1 to 8-3) is intended to help resolve such issues. The four severity levels proposed provide earlier prediction of the final outcome of frostbite injury by using a technetium-99m (99mTc) bone scan in conjunction with the clinical findings on presentation. It was noted that the probability of bone amputation for the hand and foot could also be correlated to the extent of frostbite injury seen at the time of initial presentation and early 99mTc bone scanning. In a review of 70 cases of severe frostbite injury, the probability of bone amputation was 1% for the distal phalanx, 31% for involvement of the middle phalanx, 67% for the proximal phalanx, and 98% and 100%, respectively, for involvement of the metacarpal/metatarsal and carpal/tarsal bones (see Table 8-3). Grade 1 lesions do not require hospitalization or bone scans. Grade 2 lesions may require brief hospitalization and bone scans. Rapid rewarming and treatment with antibiotics and oral vasodilators appear to be sufficient for healing. Grade 3 lesions are connected with a significant risk for amputation and require rapid rewarming, antibiotics, aspirin, and intravenous (IV) vasodilators. Grade 4 lesions have high risk for amputation and complications such as thrombosis, sepsis, and other systemic problems (see Table 8-3). Cauchy’s clinical/99mTc-based classification scheme appears particularly useful in its ability to predict at a very early stage subsequent outcome from a frostbite injury.
|Extent (Level of Involvement)||Probability of Bone Amputation (95% CI)|
|3 (proximal phalanx)||83 (66;100)|
|2 (intermediary phalanx)||39 (25;52)|
|1 (distal phalanx)||1 (00;03)|
|4 (metacarpal/metatarsal)||98 (93;100)|
|3 (proximal phalanx)||60 (45;74)|
|2 (intermediary phalanx)||23 (10;35)|
|1 (distal phalanx)||0—|
|Hand and foot||5 (carpal/tarsal)||100—|
|4 (metacarpal/metatarsal)||98 (95;100)|
|3 (proximal phalanx)||67 (55;79)|
|2 (intermediary phalanx)||31 (22;41)|
|1 (distal phalanx)||1 (00;02)|
From Cauchy E, Chetaille E, Marchand V, et al: Retrospective study of 70 cases of severe frostbite lesions: A proposed new classification scheme, Wilderness Environ Med 12:248, 2001. CI, Confidence interval.
Air alone is a poor thermal conductor, and cold air alone is not nearly as dangerous a freezing factor as is a combination of wind and cold.194 Wind velocity in combination with temperature establishes the windchill index. For example, an ambient temperature of −6.7° C (−19.9° F) with a 45-mph wind has the same cooling effect as does a temperature of −40° C (−40° F) with a 2-mph breeze184,186,189 (Figure 8-7, online). Thus, it is important to think in terms of heat loss, not cold gain. Frostbite occurs when the body is unable to conserve heat or protect against heat loss.
The type and duration of cold contact are the two most important factors in determining the extent of frostbite injury.184,186,189 Touching cold wood or fabric is not nearly as dangerous as is direct contact with metal, particularly by wet or even damp hands.51 This is a result of differences in thermal conductivity between the materials.
Deep, loose snow, which traditionally has been thought to insulate from the cold, may actually contribute to frostbite. Temperature measured beneath deep snow is frequently much lower than that on the surface. Washburn189 recounts one expedition to Mt McKinley, Alaska, when members of his party found it extremely difficult to keep their feet warm, despite a clear, sunny, −16° C (3.2° F) day with little wind. One member inadvertently dropped a thermometer in the snow and noted that it registered −25.6° C (−14.1° F). Feet must be dressed for the temperature at their level, not for surface temperature protection.189
Ambient temperature drops by approximately 1.0° C (1.8° F) for every 150 m (492 feet) of height gain. Many serious cases of frostbite originate at high altitude, but it is difficult to sort out the independent risk for hypobaria/hypoxia versus cold exposure. Hashmi and colleagues61 reported 1500 cases of high-altitude frostbite and observed a “very steep upward curve” beyond a height of 5182 m (17,000 feet) above sea level. CIVD has been shown to be diminished in non-native visitors to high altitude.52 Some important sequelae of high-altitude acclimatization, namely erythrocytosis (an increase in red blood cells to maintain oxygen transport in the hypobaric environment) and high-altitude dehydration, result in hyperviscosity that may make frostbite more likely. Garvey and associates50 demonstrated that erythremia in the setting of a hypobaric environment provoked procoagulability in a rhesus monkey model.
Recent evidence using a real-time video-imaging technique compared the sublingual microcirculation at sea level and at altitude.103 The study showed significant reduction in microcirculatory flow index (MFI) at high altitude (4900 m [16,076 feet]) when compared with sea level in small (<25 µm) and medium (26 to 50 µm) blood vessels (Figure 8-8). Larger vessels were not studied because of the relative paucity of their representation in the vascular bed studied. The results also showed further reduction in MFI within small and medium vessels at extreme altitude (6400 m [20,997 feet]). The very marked slowing of blood flow in the microcirculation at high altitude is easily appreciated (clips can be seen online at http://www.sdfimaging.net/dsmsealevelmicrocirc.wmv and http://www.sdfimaging.net/dsm6400metermicrocirc.wmv). Stagnant hypoxia may occur in tissues as a result of reduced microcirculatory blood flow and consequent failure of oxygen mass transfer. Furthermore, disparity of oxygen supply and demand at a microvascular level could lead to heterogeneous tissue oxygenation and cellular hypoxia.73 These preliminary data are the first evidence demonstrating clear reduction in microcirculatory blood flow at altitude and may in part explain the apparent increased incidence of frostbite at extreme altitude, because a reduction in flow will be associated with reduction in heat transfer. Further studies to assess the potential reversibility with supplemental oxygen are indicated.
FIGURE 8-8 Sublingual microcirculatory flow at sea level (A) and 4900 m (16,076 feet) (B). There is a significant reduction in microcirculatory flow index at high altitude (4900 m) when compared with baseline in small (<25 µm) and medium (26 to 50 µm) blood vessels, Caudwell Xtreme Everest Expedition.
(From Martin DS, Ince C, Goedhart P, et al: Abnormal blood flow in the sublingual microcirculation at high altitude. Eur J Appl Physiol 106:473, 2009.)
Hypoxic neurologic dysfunction is a feature among nonsurvivors at extreme altitude; failure to adequately protect extremities may contribute to the high incidence of frostbite at extreme altitude.42 Inadequate fluid intake and poor nutrition are also possible risk factors.121,142
Not all victims of frostbite are exposed to cold environments. Case reports cite toxic dermatologic effects of propane, butane, chloroethane, and liquid oxygen47,91,164,179,183 application to the skin, either intentionally, but exceeding time exposure for cryotherapy, or accidentally (Figure 8-9), as in the case of severe frostbite requiring skin grafting in a child improperly using a toilet air freshener containing propane and butane.91
(Courtesy Luanne Freer, MD.)
Careless use of ice to cool a soft tissue injury can result in frostbite,55 and patients should be instructed to place a towel or other fabric between ice packs and bare skin to prevent this complication.
The degree of inadequacy of protective clothing varies with conditions and may contribute to insufficient conservation of body heat.33 Tight-fitting clothing may produce constriction, which hinders blood circulation and lessens the benefit of heat-retaining air insulation. Wet clothing transmits heat from the body into the environment, because water is a thermal conductor superior to air by a factor of about 25.85 Clothing that transmits moisture away from the body may be protective if an outer wind-resistant layer decreases heat loss. However, this wind-resistant layer must retain the same transmission capabilities; otherwise, clothing will still become moist. Clothes that decrease the amount of surface area may decrease frostbite risk. Mittens are more protective than gloves, because gloves have a greater surface area and prevent air from circulating between fingers. Poorly fitted boots notoriously generate frostbite injuries, even when worn with excess socks (Figures 8-10 and 8-11; Figure 8-10, online).
FIGURE 8-10 Early frostbite at 7950 m (26,083 feet) hours after summiting Mt Everest. Five pairs of socks were being worn. The patient was advised to self-evacuate without rewarming frostbitten feet, Caudwell Xtreme Everest Expedition.
(Courtesy Christopher H.E. Imray, MD.)
Although up to 40% of total body heat loss can occur through exposed head and neck areas,178 a more recent study has refuted the widely held belief that the head loses proportionally more heat than does the rest of the body. In fact, any uncovered part of the body loses heat in proportion to the body surface area exposed; appropriate protection is important for all exposed skin.140
Development of frostbite does not depend simply on ambient temperature and duration of exposure. Along with windchill, humidity and wetness also predispose to frostbite. Skin wetting adds an increment of heat transfer through evaporation and causes wet skin to cool faster than dry skin.123 More important, water in the stratum corneum can terminate supercooling by triggering water crystallization not only in this layer but also in the underlying tissue. Skin wetness is therefore conducive to frostbite because it allows crystallization to terminate supercooling after approximately one-half of the exposure time required by dry skin. This substantiates the following clinical observation: “It has been found that supercooling displays itself in greater degree in skin that remains unwashed. Washing the skin encourages freezing, whereas rubbing the skin with spirit and anointing it with oil discourages it. The capacity to supercool greatly would seem to be connected with relative dryness of the horny layers of the skin. It is well known that Arctic explorers leave their skin unwashed.”96
Putting on clothes in response to cold is not a reflex but requires a conscious decision. When the ability to decide or to act is impaired, there is a risk for cold-induced injury; not surprisingly, alcohol has been implicated in up to 53% of nonmilitary frostbite cases.177 Once the injury had occurred, alcohol intake probably did not significantly alter the course of events. Barillo and associates7 experimentally demonstrated increased mortality and a detrimental effect of ethanol on tissue perfusion associated with severe murine frostbite. Alcohol consumption promotes peripheral vascular dilation and increases heat loss, making an exposed part more susceptible to frostbite.149
In one series of 20 urban frostbite patients, all had overt or covert psychiatric disease.135 This prompted a retrospective review that suggested that between 61% and 65% of victims of urban frostbite suffer from psychiatric disease, and some centers now advocate psychiatric screening in all cases of urban frostbite injury.
During World War II and the Korean conflict, clinical studies indicated that cold injuries occurred with higher frequency among soldiers in retreat.129,130,193 Fatigue and apathy also increase the incidence of cold injury. When warfare is proceeding toward defeat, or in conditions of starvation, soldiers often become indifferent to personal hygiene and clothing, and the frequency of frostbite increases.83 Overexertion increases heat loss. A large amount of body heat can be expanded by panting, and perspiration further compounds the problem of chilling. Both panting and sweating consume energy, which compounds the fatigue factor.
Impaired local circulation is a primary contributor to frostbite. Cigarette smoking causes vasoconstriction, decreased cutaneous blood flow, and tissue loss in random skin flaps.94 Reus and colleagues147 documented that smoking induced arteriolar vasoconstriction and decreased blood flow in a nude subject. Although red blood cell velocity increased, the net effect was a decrease in blood flow in the cutaneous microcirculation during and immediately after smoking. Curiously, habitual heavy smokers show higher scores on the Resistance Index of Frostbite (RIF), which correlates with lower risk for frostbite.29 Empirically, one could conclude that smoking, which induces vasoconstriction, should place one at increased risk for frostbite, and research supports this.30,38,61
Drugs known to have vasoactive properties may predispose to or compound frostbite injury. Disease states that alter tissue perfusion, such as diabetes, atherosclerosis, arteritis, and Raynaud’s disease, predispose to frostbite (Figure 8-12).34,88
FIGURE 8-12 A 54-year-old man had an acute primary episode of Raynaud’s disease after surfing for 80 minutes in water that was 21° C (69.8° F). The distinctive asymmetric pallor of the terminal phalanges of the fourth digit on the right hand persisted for 40 minutes. A sharp demarcation between the second and third phalangeal joints was evident, but no other symptoms were apparent. The pallor spontaneously resolved without cyanosis or redness of the affected area. Medical evaluation subsequently revealed no disorder known to cause secondary Raynaud’s phenomenon.
(From Bluto MJ, Norman DA: Acute episode of Raynaud’s disease, N Engl J Med 347:992, 2002.)
On the basis of clinical observations, an individual who has experienced prior cold injury is placed in a high-risk category during subsequent cold exposure.118,193 For undefined reasons, cold injury sensitizes an individual, so that subsequent cold exposure, even of a lesser degree, produces more rapid tissue damage.84 Cold-induced neuropathy may play an important role in the long-term sequelae of cold sensitivity after local cold injury. An alteration in somatosensory function was found, and this was more pronounced in lower limb injuries.4
Military studies emphasize that long periods of immobility contribute to the extent of cold injury.84,85 Motion produces body heat and improves circulation, especially in endangered limbs.
The deletion genotype (DD) for the angiotensin I–converting enzyme (ACE) has been associated with increased vascular reactivity in vivo and in vitro.18,65 Kamikomaki77 proposed that a case of frostbite in a climber with ACE DD allele was caused by genetic propensity for vasoconstriction.
Civilian clinical studies are inadequate for statistical evaluation of factors such as race and previous climatic environmental exposure.84,109,111 In a recent study, blacks were found to have difficulty generating increased metabolic rate (measured by oxygen consumption  and rectal temperature) after acute cold exposure, and researchers suggest this group may be at greater risk for cold injury.40 Military studies suggest that women30 and blacks30,52 may be more susceptible to cold injury. One author postulates that blacks are three to six times more susceptible to frostbite than are whites because blacks tend to initiate shivering at lower core temperatures and tend to have long, thin fingers and toes, as well as thin arms and legs, which do not conserve heat as efficiently as do their white counterparts. In testing, black fingers cool faster when immersed in cold water and reach a lower temperature before hunting reaction ensues.52 Individuals with type O blood52 and from warmer climatic regions in the United States tend to be more susceptible.193
An increased incidence of frostbite was reported in nearly 6000 military recruits with cold-provoked white finger syndrome and in those with hand/arm vibration in both exposure classes analyzed.38,128
In most patients, the initial clinical observation is coldness of the injured part, and more than 75% complain of numbness. The involved extremity feels clumsy or “absent” because of ischemia following intense vasoconstriction. When numbness is present initially, it is frequently followed by extreme pain (76% of patients) during rewarming. Throbbing pain begins 2 to 3 days after rewarming and continues for a variable period, even after dead tissue becomes demarcated (22 to 45 days). After about a week, the victim usually notices a residual tingling sensation, a result of ischemic neuritis, which explains why this sensation tends to persist longer than other symptoms. Severity of the injury usually defines the extent of neuropathologic damage. Because different injuries are influenced by so many environmental and individual factors, there may be a great deal of variation in symptoms. In patients without tissue loss, symptoms usually subside within 1 month, whereas in those with tissue loss, disablement may exceed 6 months. In all cases, symptoms are intensified by a warm environment. Other sensory deficits include spontaneous burning and electric current–like sensations. The burning sensation, which is frequently early in presentation, subsides within 2 to 3 weeks and is usually not present in victims with tissue loss. In victims without tissue loss, the burning sensation may resume on wearing shoes or increasing activity. The electric current–like shock is almost universal (97%) in patients with tissue loss. It usually begins 2 days after injury, lasts for about 6 weeks, and is particularly unpleasant at night. All frostbite victims experience some degree of sensory loss for at least 4 years after injury and perhaps indefinitely.
The clinical appearance of frostbite depends on how quickly the injured patient presents to care, and it is important to note that this appearance may initially be deceiving.12,129 Patients in the Alps who arrive by helicopter within minutes of their injury may present quite differently than do Himalayan climbers, who have nearly all rewarmed during self-descent and may be fortunate to arrive at definitive care even several days after injury. Regardless of venue, only a few patients arrive with tissue still frozen. At first, the extremity appears yellowish white or mottled blue. It may be insensate and may appear frozen solid, regardless of the depth of the injury. With rapid rewarming there is almost immediate hyperemia, even in some of the most severe injuries. Sensation returns after thawing and persists until blebs appear. At this point, some effort may be made to assess the severity of the injury.
After the extremity is rewarmed, edema appears within 3 hours and lasts 5 days or longer, depending on the severity of the case.129 Vesicles or bullae appear 6 to 24 hours after rapid rewarming.112,130 Clear bullae confer a better prognosis than hemorrhagic bullae, which indicate deeper injury. During the first 9 to 15 days, severely frostbitten skin forms a black, hard, and usually dry eschar, whether or not vesicles are present (Figure 8-13).130 Mummification forms an apparent line of demarcation in 22 to 45 days.129
In 1957, Hurley68 stated, “Tissue cells can be affected by freezing in three different ways: (1) a certain number of cells are killed; (2) a certain number remain unaffected; and (3) a large number are injured but may recover and survive under the right circumstances.” Clearly, the major treatment effort must be to salvage as many cells in the third group as possible. Frostbite treatment is directed separately at the pre-thaw and post-thaw intervals.
The International Commission for Alpine Rescue (ICAR) gives specific recommendations for self-rescue to those working, recreating, or otherwise exposed to a cold environment (Box 8-1). These guidelines advise seeking shelter from cold and wind, drinking warm fluids, removing wet clothing, taking ibuprofen, and attempting self-rewarming for 10 minutes. If at high altitude, supplemental oxygen is advised, and if sensation does not return, the victim is advised to discontinue any further exposure and seek treatment.172 Although these guidelines may seem common sense to most, climbers appreciate these guidelines for safety and self-rescue when in an exposed, potentially dangerous, and remote setting.
Guidelines From ICAR for Self-Treatment of Potential Frostbite
Adapted from Syme D (for ICAR Medical Commission): Position paper: On-site treatment of frostbite for mountaineers. High Alt Med Biol 3:297, 2002.
If transport will be prolonged (more than 1 to 2 hours), frostbite will often thaw spontaneously. It is more important to prevent hypothermia than to rewarm frostbite rapidly in warm water. This does not mean that a frostbitten extremity should be kept in the cold to prevent spontaneous rewarming. Anticipate that frostbitten areas will rewarm as a consequence of keeping the patient warm and protect them from refreezing at all costs.31
The Joint Commission of Health and Human Services, emergency medical services, and public health departments for Alaska have published further guidelines for prehospital and bush clinic care of frostbite. These guidelines are widely regarded as state-of-the-art recommendations and appear in Box 8-2. If a patient is referred from a nearby location, no attempt at field rewarming is indicated. Vigorous rubbing is ineffective and potentially harmful. The extremity should not be intentionally rewarmed during transport and should be protected against slow partial rewarming by keeping the patient away from intense campfires and car heaters. All constrictive and wet clothing should be replaced by dry, loose wraps or garments. The extremity is padded and splinted for protection, and ibuprofen 400 mg twice daily (which may be more beneficial than aspirin because aspirin may block more of the inflammatory cascade than is helpful) may be initiated. Although there is a correlation between the length of time tissue is frozen and the amount of time required to thaw that tissue, there is no direct correlation between the length of time tissue is frozen and subsequent tissue damage. Still, “rapid” transport of frostbite patients (within 2 hours) is appropriate. Otherwise, rapid rewarming should be instituted (goal to see blush of rewarming and/or 15 minutes immersed in rewarming fluid) and the victim transported with protective, dry, nonadherent dressings to prevent refreezing. Appropriate and adequate analgesia should be administered (opiates either IV or intramuscularly [IM]) may be required). Blisters should be left intact. Patients with long transport times are at greater risk for (refreezing) recurrent injury. All efforts should be made to prevent subsequent refreezing, because this creates an infinitely worse result than does delayed thawing (Figures 8-14 to 8-16; Figure 8-16, online). A victim who must walk through snow should do so before thawing frostbitten feet (see Figure 8-10). During transport, the extremities should be elevated and tobacco smoking prohibited.119 Alcohol ingestion is contraindicated.
Alaska State Guidelines for Prehospital Treatment of Frostbite